Gait Pattern Recognition Using Accelerometers

TL;DR

This study demonstrates high-accuracy gait pattern recognition using accelerometer data from wearable sensors, with effective feature extraction and classification methods achieving near-perfect results.

Contribution

It introduces a wireless sensor platform for gait data collection and evaluates multiple classifiers, achieving high recognition accuracy.

Findings

Decision Tree classifier achieved 99.4% accuracy.

K-Nearest Neighbors classifier achieved 100% accuracy.

Effective feature extraction from accelerometer data enhances recognition performance.

Abstract

Motion ability is one of the most important human properties, including gait as a basis of human transitional movement. Gait, as a biometric for recognizing human identities, can be non-intrusively captured signals using wearable or portable smart devices. In this study gait patterns is collected using a wireless platform of two sensors located at chest and right ankle of the subjects. Then the raw data has undergone some preprocessing methods and segmented into 5 seconds windows. Some time and frequency domain features is extracted and the performance evaluated by 5 different classifiers. Decision Tree (with all features) and K-Nearest Neighbors (with 10 selected features) classifiers reached 99.4% and 100% respectively.